Tactical Wheeled Vehicle (TWV) Fuel Economy Improvement Breakeven Analysis. Presented at SCEA/IPSA

Transcription

1 Tactical Wheeled Vehicle (TWV) Fuel Economy Improvement Breakeven Analysis Presented at SCEA/IPSA DISTRIBUTION STATEMENT A Approved for public release; distribution is unlimited. Raymond Kleinberg Operations Research/Systems Analyst

2 Agenda Background Data Analysis Output Future Use Refinements Questions

3 Background Tasked by Program Manager Tactical Vehicles (PM TV) to develop a flexible, Excel-based model to calculate breakeven times (savings/investments) for upgrading/modifying existing TWV fleet with fuel efficiency technology. Primary purpose would be to have a method to inform industry and rate/compare commercial technologies based on efficiency improvements and breakeven goals.

4 Data Initially, only road-load/full-load fuel tests. standard tests lock the transmission gear, vary speed not adequate for mission-level fuel economy Second round saw course testing Hybrid Electric Vehicle Evaluation and Analysis (HEVEA) program tested all major governmentowned tactical vehicles, across all vehicle classes (light/medium/heavy) on a variety of test courses

5 HEVEA Data Test Vehicles (17 vehicles across 6 systems) HMMWV M1113 M1152 Up-Armored HMMWV (UAH) HMMWV XM1124, Hybrid Electric HMMWV Remote Surveillance Targeting Vehicle (RSTV) (Hybrid Electric), GDLS Future Tactical Truck System (FTTS) Utility Variant (UV) - AM General, International MG, Lockheed Martin Maneuver Sustainment Vehicle (MSV), BAE Heavy Expanded Mobility Tactical Truck (HEMTT) A2 (base and UA) A3 OTC UA A4 Family of Medium Tactical Vehicles (Light Medium Tactical Vehicle (LTV) & Medium Tactical Vehicle (MTV)) M1078 (LMTV, 2.5T) Base, LSAC, and Continuous Variable Transmission (CVT) BAE M1084 (MTV, 5T) M1086 HE BAE

6 HEVEA Data Test Courses (4 terrains, road/full loads, and idling) Terrains Harford Loop Primary Roads Munson Standard Fuel Course (SFC) Secondary Roads Churchville B Cross Country Perryman 2 & 3 - Trails Road/Full Loads Road Load measures fuel consumption against rolling, wind, and grade forces Full Load also adds in a residual towed force Idling Stationary vehicle in neutral Varies engine speeds from rated idle to governed engine speed

7 HEVEA Data Example Graph of fuel economy data accompanied by regression equations for each plot as well as appropriate bounds on speed, and resulting fuel economy.

8 Analysis Converted all fitted equations to polynomial (quadratic) form some were constant/linear/quadratic did it for consistency reasons only fits were identical (99.7% R 2 vs 99.8% R 2 ) Now have a set of polynomial equations in one variable (speed) Created a surface of random points in (weight, speed) space using this set of equations created surface for each terrain (primary, secondary, etc.) best fit a power model to surface Excellent fits across all weight and speed ranges R 2 in excess of 99% CV no more than 6% Adjusted for Heavy fleet and operating over/under designed weight

9 Weight Adjustment Added in over/under weight factor to account for vehicles operating at other than ideal weights for which they were designed. Issue presented itself with the variance between the HMMWV M1113 and M1152. Basically, a HMMWV operating at two different weights 11,500 and 15,200 pounds AMSAA had initially validated (ca 2006) a linear relationship (best case scenario) from +/- 20% weight changes. Found a linear relationship on primary roads, but a substantial difference on cross-country and trails in excess of a quadratic term.

10 Analysis Summary Based on parametrics already fitting raw data. Same degree of fit (99%/6%), but would slightly change analysis from raw data. Hope to refine my equations by obtaining raw data from tests. Heavy Tactical Vehicles (HTV) were different enough from the Light and Medium fleets to require a dummy variable to separate them out. Added in factor to take over/under weight operation into account a very real occurrence, especially on the overweight side.

11 Final Equations MPG = W MPG = 1143 W MPG = W MPG = W S S S S HTV HTV HTV HTV

12 Additional Data OPTEMPO Utilized Operating & Sustainment Management Information System (OSMIS) Army O&S online database Gathered 3-year peacetime (w/o CONOPS) and wartime (ONLY CONOPS) data Used 99% of the mileage data to estimate OPTEMPO for each variant (e.g. M1152) and also for each system (e.g. all HMMWVs). Chose to use the most representative 99% of the activity (miles) to compute fleet OPTEMPOs Removed 1% of data. Considered extreme outlier data that consists of unused HMMWVs or those awaiting repairs at depots Idle fuel economy Cannot be estimated via a simple equation and must be entered as a value for each specific vehicle.

13 OPTEMPO Example Will be including distributions around these averages. Based on years, systems, MACOMS, etc., there can be a significant variance for OPTEMPO. This variance will be one source for uncertainty in the model, allowing ranges for breakeven times, not just a single year value.

14 Model Interface This is the primary view to enter user modified vehicle information such as mission profile (A/C), operating weight (B), fuel improvement % (B), and idling consumption (D). Notice that all the variables addressed earlier are identified here: weight, speed, over/under weight %, and terrain. The resulting fuel economies (base/improved) are calculated at the bottom (6.18/7.21 mpg) (E).

15 Cost Inputs This sheet allows entering cost specific information. This includes the current cost of fuel, investment/modification costs for both the vehicle and fleet, as well as a modification schedule/phasing of these costs. This is also the tab where fleet statistics, inflation indices, etc. are kept.

16 Output As mentioned earlier, there are confidence levels associated with the breakeven. The only input with variance at this point is the OPTEMPO. To read this chart, one would simply follow the desired confidence level curve until the investment curve flatlines (total investment reached). Under the assumptions on the last slide, at the 50% confidence level, the breakeven year is between

17 Output This chart shows a more traditional stoplight chart of investment strategy. When the SIR < 1, the decision to change status quo would never be made for financial reasons. When SIR = 1, you exactly break even. When SIR > 1.3, a positive decision for financial reasons would be made. Note, again, that the user can select their level of risk by choosing a confidence level. At the 50% confidence level, the breakeven would occur between , but most likely not before 2033 (SIR = 1.3).

18 Future Use Potential use of this tool is very open due to the built-in powers/features in Excel. Excel is a common platform that most analysts are comfortable using, editing, and expanding. Excel contains add-ins such as GoalSeek that will allow many questions to be answered with just a few key strokes. What must investment cost per vehicle be limited to in order to breakeven in 15 years? How much improvement can we buy for $5,000 per vehicle and still breakeven in 12 years? How much should we pay for a 12% improvement if we are required to breakeven in 8 years? Basically, GoalSeek will allow you to calculate ONE unknown automatically by setting the desired value of another.

19 Refinements Currently, the model works for a single vehicle type (one weight, one mission, etc.) Would like to expand this to include fleet mixes such as: LTV (All HMMWVs) Only LTV and MTV (HMMWVs, 2.5/5.0T trucks) Tactical Vehicles in Brigade Combat Team (BCT) All Tactical Vehicles (TV)

20 Questions? I thank you very much for your attention! Raymond Kleinberg US Army TACOM, Cost & Systems Analysis ATTN: AMSTA-CSB-V MS# East Eleven Mile Road Warren, MI (DSN 786) raymond.p.kleinberg.civ@mail.mil